How to send a transcript instead of a voice memo

Voice memos are easy to send and slow to receive. Recording one takes exactly as long as saying the thing; listening to one takes exactly as long again, and it has to happen somewhere the recipient can have sound on, which is not always where they are when the message arrives. Text has none of that friction — it can be read in a glance, in a meeting, in a language the reader can paste elsewhere to translate. That gap is the whole reason to send words instead of audio.

What you gain by sending text

• It can be read silently. No headphones, no finding a quiet corner, no playing it at 2x and hoping nothing important got mumbled.
• It can be searched and quoted. Pulling one sentence out of an audio file means scrubbing a timeline; pulling it out of text means selecting it.
• It is far smaller. A few paragraphs of text is nothing to send over any connection; several minutes of audio is not, especially somewhere with a weak signal.
• It can be pasted into a translator, a document, or a reply, none of which work directly on an audio file.

What you give up

Tone does not survive the trip. Sarcasm, hesitation, the difference between a firm no and a reluctant one — all of that lives in the audio and disappears in the transcript. If any of that is the actual point of the message, sending only text flattens it, sometimes badly enough to be misread. And a transcript is only as trustworthy as the recognizer that produced it: a name, a number or a technical term it mis-heard becomes something the recipient has no way to catch, because they never hear the original.

Fix it before you send it, not after

This is the one step worth never skipping. Read the transcript against the audio once before sending — speech recognition reliably stumbles on names, jargon and anything said quickly, and those are exactly the words a reader has no way to double-check once the text is in front of them instead of you. A transcript you have not checked is you asking the recipient to trust a guess.

How to actually send it

Once the transcript reads the way you want it to, getting it out is the same as sharing anything else on iPhone: copy the text and paste it directly into a message, or use the share sheet to send it as a file to Mail, Messages, Notes, or wherever the recipient will actually see it. Plain text is almost always the right choice here — it opens as readable words on any device, with nothing for the other person to figure out.

When the audio is still worth including

Not every case is text-only. If tone genuinely matters — a message that needs to sound reassuring rather than curt, or one where a flat "fine" and a warm "fine" mean opposite things depending on how it was said — attach the audio alongside the text rather than instead of it. The transcript then does its job as a fast preview and a searchable record, and the recording is there for the one time someone needs to actually hear it. Sending both costs nothing extra; it is one more share-sheet tap.

Doing this in Voice Studio

Any recording’s transcript can be copied straight to the clipboard or exported as a TXT file through the share sheet, which is the plain, readable format for pasting into a message or an email. The audio is right there too, as the original .m4a, if you decide the recipient should hear it rather than just read it — nothing about sending the text requires giving up the option to send the audio as well. The transcript stays editable in the app until you send it, so correcting a name the recognizer missed takes a moment before it becomes someone else’s copy of the mistake.

Common questions

Is it rude to send a transcript instead of a voice memo?

Not inherently — it usually reads as respecting the other person’s time. It becomes a problem only when tone was the actual point of the message and gets lost in the text, which is worth watching for rather than a reason to avoid text generally.

Should I send the TXT or the JSON export when sharing with someone else?

TXT. It reads as plain, formatted words on any device. JSON keeps the data in separate fields for a program to read, which shows up as clutter to a person just trying to read a message.

What if I only want to send part of a long transcript?

Copy the section you need and paste just that into a message rather than exporting and trimming a file — for a short excerpt, copying is faster than exporting the whole transcript and cutting it down afterward.

Can the recipient reply to a transcript the way they would to a voice memo?

They can reply however they normally would to a text message, which is often the point — a transcript can be answered with a quick line back, where a voice memo tends to invite another voice memo.